I had my own little Techtool vs Norton battle a litle over a year ago, I maintain 5 macs a work and 3 at home. Techtool won, I never analysed it enough to know exactly why. I just noticeds that one day i was taking Norton off all my systems. I still turn off all tha auto matic shit that comes on at startup and shutdown.
Anyone have any benchmarks that prove there is a speedup in system performance after defragging a HD? I'm talking about a modern, 5400 or 7200 RPM HD, at least an ATA66 interface.
I doubt you'll find any benchmarks, because the the difference in performance between a moderately fragmented HD and an "optimized" HD is negligible.
But I'd love to be proven wrong...it would give me something to do with Norton and TTP that I spent all kind of money on.
I think it depends on which tool you're using to optimize the disk but I've noticed with Techtool (on my OS 9 disks) that boot times are a little quicker and that certain applictions (whichever get placed in the sweet spot on the drive I guess) are a little quicker to boot / open / save files. But it is a small difference. Maybe a second or two in most instances.
One other thing optimization might do is decrease the liklihood of corrupted files? Not asserting, just taking an educated guess.
<strong>P.S. DiskWarrior is great at what it does. Doesn't do much, but it does it well. I always run it first.</strong><hr></blockquote>
Couldn't agree more. With the exception of one occassion (which required the return of the iBook for a hard disk replacement) DiskWarrior has recovered non-booting Macs at least five or six times for me.
I've never had a problem with a Mac that needed a more "sophisticated" app.
Most of the other stuff included in monitor/repair/recover apps seems utterly pointless.
<strong>Here's a question for someone in the know.
Is XP any better at fragmentation than previous versions? We still use 95 in our dept. at work (yuk ). <img src="graemlins/oyvey.gif" border="0" alt="[No]" /> </strong><hr></blockquote>
There have been studies showing that a server left to its own design will lose a great deal of its performance thus impacting the bottom line in terms of speed and cost. This <a href="http://www.execsoft.com/diskeeper/idc-white-paper.pdf"; target="_blank">White Paper</a> actually goes into just how much fragmentation affects performance on NT/2000.
OS X is just another OS that uses a disk. It is not immune to the effects of Fragmentation. It may however do more to prevent fragmantation of individual files in the first place.
I've experienced Norton saving my HD from being wiped over the last 9 years when it was unbootable SEVERAL times. I'd hardly call that " utterly pointless ".
I've experienced Norton saving my HD from being wiped over the last 9 years when it was unbootable SEVERAL times. I'd hardly call that " utterly pointless ".</strong><hr></blockquote>
Yikes! No need to be so testy. I didn't say Norton was utterly pointless. See:
[quote]<strong>Most of the other stuff included in monitor/repair/recover apps seems utterly pointless.</strong><hr></blockquote>
i'm not very hd savvy.. but i will add this, it makes sense that optimization (bunching together frequently used files at the front and moving frequently changing ones at the end or middle) would do wonders for os x simply because when i placed my swap file(s) on their own dedicated partition os x sped up noticeably, xlr8 i believe did a work up on it, in 10.1 i went from 3-4 bounces to launch system panels to 1-2 :eek: in fact i cut launch time by at least half and usually over on most apps- if it didn't speed up it didn't slow down, so i only saw improvements- the system felt snappier
my logic is that the drive always knew right where to go to find my swap file, cutting down seek time and improving performance- therefore it makes sense that optimization would make life easier for the drive by cutting down seek time, even in os x, and improving performance...
The thing that junkyard doesn't get is that even if hard drives are faster, files and hard drives are larger. Also in the age of new and improved OS's the data on the hd get's updated more often. Back in the system 7 era things could literally go for years without changing much.
Sorry Belle, didn't mean to get testy I'm home sick from work today so I'm a little cranky.
<strong>Here's a question for someone in the know.
Is XP any better at fragmentation than previous versions? We still use 95 in our dept. at work (yuk ). <img src="graemlins/oyvey.gif" border="0" alt="[No]" /> </strong><hr></blockquote>
It sure is, if you format the drive in NTFS format, not FAT32.
<strong>First of all, there is no magic in OSX. just because it's a new os, it's still HFS+, and it becomes fragmented just like every other fs (except for something like ffs, which is nice when coupled with softupdates)
</strong><hr></blockquote>
Not sure you actually meant to imply this, but SoftUpdates don't have any impact on fragmentation (but they would nonetheless be a welcome addition to OS X's UFS implemntation).
<strong>Anyone have any benchmarks that prove there is a speedup in system performance after defragging a HD? I'm talking about a modern, 5400 or 7200 RPM HD, at least an ATA66 interface.
</strong><hr></blockquote>
As has been stated before, this heavily depends on how fragmented the hard drive was in the first place (and thus on the way you use your machine).
While I agree that MS' claim that NTFS was immune to fragmentation issues has been proven wrong (and MS' quietly admitted that by including a defrag utility with Win2k and XP), it is, at least from my experience, still much better than FAT(32).
<strong>when i placed my swap file(s) on their own dedicated partition os x sped up noticeably
(...)
my logic is that the drive always knew right where to go to find my swap file, cutting down seek time and improving performance
</strong><hr></blockquote>
This is highly unlikely, unless that other partition was on another drive.
The drive itself doesn't understand the concept of partitions or files, all it knows about is how to write / read raw blocks to / from disk.
Partitions, files, etc. are all handled in the lower layers of the OS.
Thus, for knowing where a certain file is on disk, it doesn't make a difference from the drive's point of view whether the swap file is on it's own partition or not.
<strong>A disk controller may even write an updated file in the open space then delete the reference to the original even if it does not change size, that's up to the manufacturer.
</strong><hr></blockquote>
This is technically incorrect.
All the drive controller does is reading / writing raw blocks from / to the disk. It doesn't even know what a file or directory is.
Thus, the stuff you described above would happen not inside the drive controller, but inside the OS' file system implementation.
<strong>When you install an OSX update it prebinds so without getting fancy is is pretty much doing the same thing as diskwarrior.
</strong><hr></blockquote>
Where did you get this from? Prebinding has nothing to do with disk fragmentation, and defragmenting your drive doesn't do any harm to prebinding.
What prebinding basically is about is calculating target addresses for jumps from an app's code to code inside shared libraries. This is usually done at run-time, but prebinding can do some of it in advance, thus speeding up program launch. (This is also the reason why prebinding has to be re-done whenever the system frameworks change.)
Comments
I doubt you'll find any benchmarks, because the the difference in performance between a moderately fragmented HD and an "optimized" HD is negligible.
But I'd love to be proven wrong...it would give me something to do with Norton and TTP that I spent all kind of money on.
One other thing optimization might do is decrease the liklihood of corrupted files? Not asserting, just taking an educated guess.
When I run Norton it often says something like "system" bundlebit is on, but should be off. So I change it.
Then I run TechTool (a superior, if slightly slower app IMO) and it says "system" bundlebit is off, but should be on.
Which is right? And why?
And what the heck is a bundle bit?
And why do fonts always end up with "bad modification dates" even if they haven't been modified?
P.S. DiskWarrior is great at what it does. Doesn't do much, but it does it well. I always run it first.
[ 02-06-2002: Message edited by: seb ]</p>
<strong>P.S. DiskWarrior is great at what it does. Doesn't do much, but it does it well. I always run it first.</strong><hr></blockquote>
Couldn't agree more. With the exception of one occassion (which required the return of the iBook for a hard disk replacement) DiskWarrior has recovered non-booting Macs at least five or six times for me.
I've never had a problem with a Mac that needed a more "sophisticated" app.
Most of the other stuff included in monitor/repair/recover apps seems utterly pointless.
<strong>Here's a question for someone in the know.
Is XP any better at fragmentation than previous versions? We still use 95 in our dept. at work (yuk ). <img src="graemlins/oyvey.gif" border="0" alt="[No]" /> </strong><hr></blockquote>
What I have noticed being the system admin here is the if the computer is using NTFS it actually seems to fragment WORSE than FAT or FAT32. Microsoft made many claims that NTFS was frag proof, but time did not bear out their conclusions. Even now on their web sites they have posted <a href="http://www.microsoft.com/technet/treeview/default.asp?url=/TechNet/prodtechnol/windows2000serv/maintain/optimize/w2kexec.asp=/Tech"; target="_blank">This </a>.
There have been studies showing that a server left to its own design will lose a great deal of its performance thus impacting the bottom line in terms of speed and cost. This <a href="http://www.execsoft.com/diskeeper/idc-white-paper.pdf"; target="_blank">White Paper</a> actually goes into just how much fragmentation affects performance on NT/2000.
OS X is just another OS that uses a disk. It is not immune to the effects of Fragmentation. It may however do more to prevent fragmantation of individual files in the first place.
I've experienced Norton saving my HD from being wiped over the last 9 years when it was unbootable SEVERAL times. I'd hardly call that " utterly pointless ".
<strong>Belle,
I've experienced Norton saving my HD from being wiped over the last 9 years when it was unbootable SEVERAL times. I'd hardly call that " utterly pointless ".</strong><hr></blockquote>
Yikes! No need to be so testy. I didn't say Norton was utterly pointless. See:
[quote]<strong>Most of the other stuff included in monitor/repair/recover apps seems utterly pointless.</strong><hr></blockquote>
my logic is that the drive always knew right where to go to find my swap file, cutting down seek time and improving performance- therefore it makes sense that optimization would make life easier for the drive by cutting down seek time, even in os x, and improving performance...
YMMV
Sorry Belle, didn't mean to get testy I'm home sick from work today so I'm a little cranky.
<strong>Here's a question for someone in the know.
Is XP any better at fragmentation than previous versions? We still use 95 in our dept. at work (yuk ). <img src="graemlins/oyvey.gif" border="0" alt="[No]" /> </strong><hr></blockquote>
It sure is, if you format the drive in NTFS format, not FAT32.
(The same is true for WinNT and Win2k).
Bye,
RazzFazz
<strong>First of all, there is no magic in OSX. just because it's a new os, it's still HFS+, and it becomes fragmented just like every other fs (except for something like ffs, which is nice when coupled with softupdates)
</strong><hr></blockquote>
Not sure you actually meant to imply this, but SoftUpdates don't have any impact on fragmentation (but they would nonetheless be a welcome addition to OS X's UFS implemntation).
Bye,
RazzFazz
<strong>Anyone have any benchmarks that prove there is a speedup in system performance after defragging a HD? I'm talking about a modern, 5400 or 7200 RPM HD, at least an ATA66 interface.
</strong><hr></blockquote>
As has been stated before, this heavily depends on how fragmented the hard drive was in the first place (and thus on the way you use your machine).
Bye,
RazzFazz
<strong>
What I have noticed being the system admin here is the if the computer is using NTFS it actually seems to fragment WORSE than FAT or FAT32. Microsoft made many claims that NTFS was frag proof, but time did not bear out their conclusions. Even now on their web sites they have posted <a href="http://www.microsoft.com/technet/treeview/default.asp?url=/TechNet/prodtechnol/windows2000serv/maintain/optimize/w2kexec.asp=/Tech"; target="_blank">This </a>.
</strong><hr></blockquote>
While I agree that MS' claim that NTFS was immune to fragmentation issues has been proven wrong (and MS' quietly admitted that by including a defrag utility with Win2k and XP), it is, at least from my experience, still much better than FAT(32).
Bye,
RazzFazz
<strong>when i placed my swap file(s) on their own dedicated partition os x sped up noticeably
(...)
my logic is that the drive always knew right where to go to find my swap file, cutting down seek time and improving performance
</strong><hr></blockquote>
This is highly unlikely, unless that other partition was on another drive.
The drive itself doesn't understand the concept of partitions or files, all it knows about is how to write / read raw blocks to / from disk.
Partitions, files, etc. are all handled in the lower layers of the OS.
Thus, for knowing where a certain file is on disk, it doesn't make a difference from the drive's point of view whether the swap file is on it's own partition or not.
Bye,
RazzFazz
<strong>A disk controller may even write an updated file in the open space then delete the reference to the original even if it does not change size, that's up to the manufacturer.
</strong><hr></blockquote>
This is technically incorrect.
All the drive controller does is reading / writing raw blocks from / to the disk. It doesn't even know what a file or directory is.
Thus, the stuff you described above would happen not inside the drive controller, but inside the OS' file system implementation.
Bye,
RazzFazz
<strong>When you install an OSX update it prebinds so without getting fancy is is pretty much doing the same thing as diskwarrior.
</strong><hr></blockquote>
Where did you get this from? Prebinding has nothing to do with disk fragmentation, and defragmenting your drive doesn't do any harm to prebinding.
What prebinding basically is about is calculating target addresses for jumps from an app's code to code inside shared libraries. This is usually done at run-time, but prebinding can do some of it in advance, thus speeding up program launch. (This is also the reason why prebinding has to be re-done whenever the system frameworks change.)
Bye,
RazzFazz
<strong>I'm actually kinda curious about this.
i know that the various file systems makes fragmentation occur less (such as freebsd's fs, which i forget it's name right now)
</strong><hr></blockquote>
Actually, FreeBSD uses UFS too (a newer implementation though, and adds some nice additional goodies, such as SoftUpdates).
Bye,
RazzFazz